Recognition Algorithm Polymorphic Palm

With the rapid development of society and technology, people pay more attention to identity authentication of accuracy, practicability, security, etc. However, the traditional identification methods can’t meet realistic demand. Therefore, a large number of scholars have been searching for a safe and reliable method. Biometrics refers to the study of methods for recognizing humans based on one or more physical or behavioral traits, which is substituted for the traditional identification method gradually.Palmprint recognition is a novel biometric recognition. Compared with other biometric identification technology, palmprint has more abundant texture information and direction information. Together with its uniqueness and lifelong invariance, palmprint become a kind of high potential and effective identification method, attracting more and more scholars enter into this field.Palmprint information have difference in every direction, which hold different weight in every scale and affect recognition rate. So fusing the unique characteristic of palm and recognizing correctly have been developed. In order to solve the above problem, we proposed a novel palmprint recognition algorithm based on polymorphism roughly from four aspects:the standpoint of first part is the sharp and direction of palmprint. In order to get lower dimension and identify correctly, we contrast the energy after Gabor transform in two angle, and except unnecessary image.The second part we seriously survey palmprint ridge, analyzing the obvious palmprint ridge and obtaining its local information intensity. Lastly we blend them in parallel combining section thinking. In the last aspect, we try to another algorithm called Curvelet transform which can substitute Gabor transform. This algorithm can describe the scale,direction of palmprint ridge, ensure precision, identify faster than Gabor transform. So we propose a palmprint recognition algorithm based on CLBP. By contrasting every layer of Curvelet transform, we choose some of them which contain almost palmprint character, and blend Local Binary Pattern to identify palmprint. Finally, in view of the problem of feature extraction difficult, poor stability and large characteristic dimension, we further study Curvelet transform and propose a novel palmprint recognition algorithm based on spatial matching. We can match different palmprint by treating invariable feature as character, which is a simple, stable and rapid algorithm.